AssemblyAI was founded in 2017 by Dylan Fox in San Francisco. The company provides developer-friendly APIs for automatic speech recognition (ASR), speaker diarization, content moderation, and audio summarization.
AssemblyAI raised over $115 million, including a $50 million Series C in 2023. Investors include Insight Partners, Accel, and Y Combinator. The funding supports development of their proprietary speech models, which compete with offerings from Google, AWS, and OpenAI.
The company’s Universal-1 model, released in 2024, achieves near-human accuracy on English speech recognition and supports multiple languages. Beyond transcription, AssemblyAI provides LeMUR — an LLM framework built specifically for audio data that can answer questions about, summarize, and extract action items from transcribed conversations.
AssemblyAI processes billions of minutes of audio annually for over 200,000 developers. Use cases include meeting transcription, podcast indexing, call center analytics, and accessibility tools. The API handles real-time streaming and batch processing.
The company differentiates from big cloud ASR services with better accuracy on challenging audio (accents, background noise, domain-specific terminology) and a developer experience focused entirely on speech and audio intelligence. AssemblyAI employs around 130 people and maintains its position as a leading independent speech AI provider.
